Supermicro liquid cooling solutions can reduce OPEX by up to 40%, and allow data centers to run more efficiently with lower PUE. Supermicro has proven liquid cooling deployments at scale and enables data centers operators to deploy the latest and most performant CPUs and GPUs.
One of the most powerful liquid-cooled AI SuperClusters is designed to take xAI’s Grok AI to the next level. With extensive experience deploying large-scale direct-to-chip (DLC) liquid-cooled AI systems, Supermicro is developing leading liquid cooling technology powering the world’s most ambitious AI infrastructure projects.
Linus heads to the roof to see how Supermicro is transforming data center efficiency with their complete liquid cooling solutions, offering everything from cold plates to cooling towers, with capacity to deploy over 1500 liquid-cooled racks per month.
Join Level1Techs for this tour of a Supermicro 5MW custom loop data center and cooling tower. Inside, they look at how Supermicro integrates rack-scale liquid cooling solutions to support high-density GPU deployments at up to 80kW per rack.
Patrick Kennedy from ServeTheHome visits Supermicro’s Silicon Valley headquarters to take a look at some of its new state-of-the-art liquid cooled GPU systems to understand why traditional air cooling just isn’t enough for modern AI servers. While he’s there, he also takes a look at some of the key components developed by Supermicro that enable complete end-to-end liquid cooling solutions.
Supermicro is an end-to-end solution provider for rack scale liquid cooling solutions. Supermicro’s delivers a complete and fully tested solution, including servers, racks, networking and liquid cooling infrastructure.
Supermicro's in-house developed CDUs feature 1+1 redundant hot-swappable pumps that circulate coolant to cold plates, removing heat from CPUs, GPUs, and DIMMs. CDUs have a cooling capacity of 100kW or 250kW (depending on the model), enabling extremely high rack densities.
Cold plates are mounted on top of CPUs, GPUs, and DIMMs to cool components efficiently as coolant flows through micro-channels in the cold plate.
CDMs distribute coolant to each server in the rack and return the hot coolant back to the CDU.
Hose kits are designed to easily connect liquid cooling racks directly to data center primary water supply or cooling towers. The kits ensure seamless integration with existing facility piping and provides the flexibility for future expansion, potentially lowering data center upfront investment costs.
The full turn-key AI data center solution for mission-critical enterprise use cases, eliminating the complexity of building a large cluster
Accelerated time-to-delivery of plug-and-play racks with proven quality and reliability
The CDU control system optimizes power consumption while ensuring efficient cooling is delivered to all CPUs, GPUs, and DIMMs, with WebUI access and integrated into Supermicro’s SuperCloud Composer management software.
Lifecycle management of liquid cooling components, GPU/
